ON THE USE OF THE HYDROMETER OF TOTAL IMMERSION. 259 



The mode of procedure that I have followed in the deter- 

 minations themselves is as follows. 



After having opened the bottle containing the water sample 

 with all possible care, and well wiped the neck of the bottle to 

 get rid of possible grains of salt, I rinse out the cyhnder twice 

 with about 20 to 30 cub.cm. of the sample that is to be deter- 

 mined.^ 



I then fill the cylinder with sufficient water to give the 

 hydrometer plenty of room, and place it in the bath. I remove 

 all the air-bubbles on the inside of the cylinder by the aid of 

 the stirrer. 



I use two thermometers, and place them close together^ the 

 one with its bulb near the bottom, the other with its bulb near 

 the surface 2. (See PI. 23, figs. 1 and 2). 



Finally, I place the hydrometer in the sample. If any small 

 air-bubbles attach themselves to the bottom of it, they can be 

 easily removed with the stirrer. 



The platinum spiral is placed round the stem of the hydro- 

 meter (see PI. 23, fig. 3). On its left arm with one bend, I 

 hang weights down to O'l gr. The leighter weights, from 0*05 

 gr. downwards, I thread with the tweezers on to the right arm 

 inside the double bend. The object of the double bend is to 

 prevent these small weights from being left lying on the surface 

 film as they are liable to be when the hydrometer is immersed. 



' In order to prevent the formation of small air-bubbles during the obser- 

 vation, the samples are some time before the determination heated to 

 about 10° above the ordinary temperature of the room, and then well 

 shaken. They are then opened with the same care as mentioned above, 

 and are left open for about 5 minutes, to allow as much air as possible 

 to escape. The bottles are thereupon, closed. If this method be pro- 

 perly carried out, no trouble is ever experienced from air-bubbles. 

 With a uniform treatment also, all the samples that are determined 

 contain very much the same volume of gas. 



^ This is greatly simplified by having the 2 thermometers welded to- 

 gether to one stem. The distance between the bulbs of the thermo- 

 meters is proportioned to the height of the cylinder. 



